This project is about making public chemical hazard information more usable, by applying limited labor resources and Python.

• Transforming certain useful resources, such as hazard classification lists, into simple & uniform data formats.
• Helping to link different chemical identifiers, especially authority-controlled IDs & open structure-based IDs.

Many public sources of chemical hazard information don't have the kind of programmatic accessibility that's generally expected from open data. For the major sources of open chemical data on the web, there are helpful Python interfaces: see PubChemPy, ChemSpiPy, CIRpy, BioServices. But those resources still leave gaps in the accessibility of certain types of information, which I'm interested in.

• chemex package, just a loose collection of convenient code.
• scripts/ contains a few scripts that help transform and clean data from useful public-domain sources. See the README(s) in that directory.
  • For example, GHS classifications.
• notebooks/ contains notebooks for explanation and/or testing.

• Python 3.x, or 2.7 with future
• requests
• beautifulsoup4
• lxml
• pandas & numpy
• xlrd
• boltons

Files in data/ are from the public domain.

Everything else here is free and unencumbered software released into the public domain (see LICENSE).